The evolution of artificial intelligence reaches a new pinnacle with the introduction of GPT-4 Turbo by OpenAI. This advanced language model promises to transform the way we interact with technology, thanks to its extended memory, low cost and up-to-date knowledge. So here's all the news Announcements during the OpenAI DevDay held yesterday in San Francisco.
Topics of this article:
OpenAI announces GPT-4 Turbo DevDay: what's new
During OpenAI's first developer event, DevDay 2023, held on November 6 in San Francisco, company CEO Sam Altman unveiled GPT-4 Turbo. The new language model is capable of developing a much larger amount of text than its predecessor, with an extended context window and knowledge update until April 2023. Furthermore, they have been APIs for DALL-E 3 and GPT-4 Vision introduced, as well as an API for text-to-speech, marking a significant step forward in the field of assistive artificial intelligence.
Altman highlighted OpenAI's AI impact on the world, highlighting its role in technological accessibility. He shared some impressive data: further 2 million developers use OpenAI APIs to build applications, and more than 92% of Fortune 500 companies rely on their platform. ChatGPT boasts over 100 million weekly active users, a number that demonstrates the growing relevance of this technology.
The surprise of the event was the appearance of the CEO of Microsoft, Satya Nadella, who discussed with Altman the strengthening of collaboration between Microsoft and OpenAI, sharing visions about the future of technology and its potential to empower people.
Read also: GPT-4: Gemini will be Google's rival. Here are the differences
Main Features of GPT-4 Turbo
The introduction of GPT-4 Turbo marks a significant shift not only in terms of technical capabilities, but also in terms of affordability. Implementing this advanced model as an API results in significant cost savings: a third less for input tokens, which translates into substantial savings for developers who power their applications with large volumes of data.
For output tokens, the cost is halved, encouraging the use of elaborate and detailed responses in AI-based interactions. This competitive pricing strategy is also extended to GPT-3.5 Turbo's API models, demonstrating OpenAI's commitment to making its more accessible technology.
The updated token per minute limits represent another major breakthrough, doubling the amount of data that paying customers can process in real time. This expansion of limits means that applications can now work with greater fluidity and without interruptions, key elements for an optimal user experience.
New capabilities through APIs
The advancement of OpenAI's APIs allows developers to leverage new processing and interaction capabilities. With GPT-4 Turbo, it's now possible analyze images within conversations, expanding the possibilities for interaction and use cases. AI image synthesis via DALL-E 3 opens up new horizons in visual content creation, while the text-to-speech model enriches applications with realistic and natural voices, improving accessibility and user engagement.
What's next on ChatGPT
It is worth mentioning that OpenAI will enrich ChatGPT with the following innovations:
- Maker feature, which could allow users to create a GPT custom. This new feature represents a significant evolutionary leap, as it opens the doors to unprecedented customization in the world of virtual assistants
- connect ChatGPT to your Google Drive or Microsoft 365 accounts directly from the platform settings. This update represents a significant step forward towards greater integration between ChatGPT and popular digital workplaces
- new premium plan intended for teams, which includes unlimited access to GPT-4, advanced data analysis and a context 4 times longer.
